Standard for American Carhonifcrons. — Keyes 301 
The series of the system appears to be more sharply defined 
than anywhere else. Few sections have the base so abruptly 
cut off from the Devonian below. At the top the Cretaceous 
strata often rest upon it in marked unconformity. 
Instead of trying- to make the Mississippi valley section 
fit that of Pennsylvania, as heretofore has always been done, 
effort would be far more fruitful of satisfactory results by 
bringing the rock succession of the latter into accord with the 
former. In place of attempting to extend the Mississippi 
sequence to the region of the Rockies, the formations of the 
latter should be apposed to those of the first mentioned. In 
other words, a direct and complete reversal of present 
methods of comparative procedure is recommended as the 
first step in the solution of the intricate problems of correla- 
tion presented. 
The critical criteria for bringing the general sec'tions of the 
various provinces into harmony with a general schematic sec- 
tion for the continent must be very different from those which 
have been long in use. The fossils have not only failed to give 
exact results, but it is now known that in many cases their 
several elements actually admit oi no logical camparisons. 
The comiparisons of the floras have been more satisfactory, 
so far as they have gone, but they have been necessarily lim- 
ited ; and as yet there have been no means of checking these 
readings. 
A review of the work of the last fifty years clearly dem- 
onstrates that in correlating strata, the testimony of the fossils 
does not give the results expected of them. For exact cor- 
relation, we must turn to other standards. In the physical his- 
tory of a region the most hopeful outlook is presented. 
In comparing faunas which are marine in one province or 
region, brackish in another, littoral elsewhere, there are few 
or no common factors that can be readily utilized. In order 
that the real significance of the varying features may be un- 
derstood, the fossils demand vastly more study than can or- 
dinarily be given them in a short time. The sought for infor- 
mation is therefore usually obtained from other sources long 
before the fossils are able to yield it up. The mlethods of 
physical correlation of strata have now become so advanced 
that the results are highly satisfactory and exact to a remark- 
